Re: Cluster metadata (bulk indexing terabytes of time-based data)


(Berkay Mollamustafaoglu-2) #1

No, what I'm saying is that you can have multiple EBS volumes attached to
the EC2 node, and stripe them to have a single volume. This is often done
to for increased I/O.
So you'd run a single ES instance on the server, put all the data on the
striped EBS volume you've created. If the node fails, you have the data on
EBS, you can start a new EC2 instance, attach the EBS volumes, and you'd be
good to go.

By the way, I think this email jumped threads :slight_smile:

Regards,
Berkay Mollamustafaoglu
mberkay on yahoo, google and skype

On Fri, Jan 20, 2012 at 3:32 PM, Derrick derrickrburns@gmail.com wrote:

To be clear, are you suggesting one EBS volume per elasticsearch index? I
could end up with hundreds of small volumes, correct? Wouldn't I need to
run one ES node per volume?


(system) #2